Catapa Resume Parser: End to End Indonesian Resume Extraction

Abstract

This paper proposes a method to solve the problem of extracting contents from a resume, especially for Indonesian resumes using segmentation method by header followed by models for each corresponding headers. An end to end resume extraction system is created using some heuristic rules and machine learning algorithms to solve the problem. On average, an accuracy of ~91.41% is achieved for personal information entities (name, email, phone, gender, date of birth, and religion), ~68.47% accuracy for job experiences entities (company, job title, start date, and end date), and ~80.85% accuracy for educations entities (institution, major, level, start date, end date, and GPA) out of 221 random resumes using the aforementioned method.